Topic: Please do not change MAX_BLOCK_SIZE - page 14. (Read 13023 times)

I think that you miss the point, if you are having small transactions, they SHOULD NOT be audited by everyone in the world (or at least those who use Bitcoin).  The time-importance of small transactions is much lower than of high-powered-money transactions.

For-example, I could make a merge-mined chain that works like Bitcoin, where the value of any coins older 6 weeks is halved. (and forgotten once they get smaller than the smaller unit).  New miners would sell these coins, and people could trade over medium time periods.  (purchasing and selling this temporary currency for bitcoin).  However this is just one option that I made up then.  There is a multitude of possible merge-mined coins that could be developed.

However the key point, is that the whole world should not care about my transaction of going to the corner store and buying a coke.  Only transactions that are significant for everyone should be audited by everyone.  7tps seems like a good rate.


Open Transactions works completely differently to Bitcoin, they are in-fact extremely complementary.  They solve different problems.  Bitcoin solves the 'low-trust, everyone audits, value transfer.'  Open Transaction solves the problem 'anything that is related to something in the world should be able to be traded'.

It does the 'fat cat banksters' a lot more good to see your transactions than it will you to see theirs.  If they have their way, you'll definitely have most of your activity in the block chain, and there will be a fair amount of information logged about how it got there.  So, don't worry, be happy.

Ya, ya, I know.  'Tor'.  Yawn.

If it truly is win-win then you don't need an artificial limit of 7 transactions per second included in the Bitcoin protocol.

If those merge mined chains, and the off-chain transaction systems such as the ones you develop, really are valuable on their own merits then users will happily adopt them no matter how many transactions per second the Bitcoin network is capable of processing.

Open Transactions is valuable on its own merits, right? It's not that the only way you'll ever get a significant user base is if Bitcoin is artificially limited to a cripplingly low transaction rate, is it?

Fully agree. So the way forward seems to lie in two separate but significant Bitcoin enhancement projects:

1. Lightweight nodes with a pruned blockchain consisting of unspent transaction outputs (utxo) as described by etotheipi, which maaku is working to prototype. This benefits all non-mining nodes.

2. Block header propagation, perhaps requiring a pool of verification nodes. This benefits miners, reducing the need for mining pools and large block propagation.  A major technical problem exists in how this model of block-sharing copes with reorganizations. Is this solvable??

If these were implemented in a future release then Bitcoin would be well on the path for large-scale capacity while still retaining significant decentralization and low-bandwidth nodes.


If 99% of the transactions I want to do are priced away from the blockchain by high fees, then I certainly will be switching off my node rather than leaving it running to support fat-cat banksters who are monopolizing the blockchain.

+1 Very well explained.


Bitcoin is supposed to be a "digital cash". If people are willing to pay a $100 transaction fee to make a transfer, they can just use a wire transfer.

If Bitcoin becomes popular, then people will not mind paying eqv. $100 in fees for buying a house.  When a house is 0.01 BTC.

7tpc is enough.  Let Bitcoin be the high-powered-money.  Not play money that you use at the corner store.

We can create any number of merge-mined chains for any number of other transactions.  These chains can be optimized for the particular purpose that they fulfill.  They can even be dominated in BTC (with clever escrow-bitcoin-look-up).


I for one would love to only have 7tpc. This would mean that it would be easy for everyone to audit the high powered money.  (we would know what the bankers are doing with our cash).  In-fact in 10 years, my mobile phone may be able to audit the entire chain. (with some dedicated cypto hardware).


On another note, merge-mined chains that fulfill other purposes will make the main Bitcoin chain more secure. So it is a win-win.

The tradeoff will be worth it for certain scenarios--currently you already submit much of your personal data to exchange operators, with OT it's possible to provide a level of anonymity that even if government agents manage to obtain the transaction records, they can turn up hardly anything worthwhile.

Quoted... because it deserves to be said again. Thank you, justusranvier.

You're preaching to the choir

The solution is the Bitcoin blockchain.

All of these off blockchain ideas are all trying to solve the problem that Bitcoin solves, without using the amount of resources that a solution requires. It's never going to work out that way. If somebody does find a way to solve the decentralized transaction ordering problem in a way that's better than Bitcoin that technology isn't going to be an add-on to Bitcoin - it's going to be a new currency that replaces Bitcoin.

All of this effort around designing alternatives to the blockchain would be more effectively employed making the blockchain work. It's not going to be any easier to scale an off chain transaction processing system to huge rates and keep the same level of security and censorship resistance than it would be to scale the blockchain to the same level.

That's correct.  The bitcoin-denominated tokens issued on OT servers are promises issued by the aforementioned voting pool to redeem for real bitcoins at par.  OT deals with the issue of trust by spreading it out across multiple parties via a multisignature transaction on the blockchain.  What makes me uneasy is there are technical, logistical, and cognitive limits on the total number of parties able to participate in the voting pool, and so the issue of trust ultimately still remains.

Edit: Furthermore, once trust enters the equation, anonymity usually leaves, since most people aren't willing to trust their money to strangers.  Thus, we likely haven't actually attained censorship resistance.

It's not the best analogy, a Chaumian token is more similar a a banknote.

blind signatures should protect from this. as an analogy, the banks wont be moving your money around directly but rather moving around a locked box full of money that you have given them that only you know the combination to. This way the money that you give to the bank will only be valuable to you and can never be used by the bank.

If that's the end game, then I can't help but feel uneasy about having a dozen or so people collectively (via the proposed voting pool) having ultimate control over most of the bitcoins.  Systems that rely on trust tend not to be very decentralized (for cognitive reasons?).

hopefully open transactions will solve all these problems much more elegantly.

There's no particular need to have different exchange rates for different chains. What we're talking about is just sharding - it's a common way of growing systems that run into some natural scaling limitation. Without doing anything particularly clever:

When you got to a particular size you'd split the existing outputs into two shards, which would split in turn when they got too big. So some of your coins would be on Shard 1 and some would be on Shard 2. Your client would have to be able to connect to all the shards you had coins on. Most people would use SPV or whatever, but if they wanted to they could run full nodes for some, all or none of the shards.

For most purposes this would look identical to the end user to what they see at the moment. Just as your client currently looks at the amount of money you want to spend and figures out the appropriate collection of spendable outputs, it would gather together enough outputs on the same shard to make your payment.

The hitch is that if you didn't have enough coins on a single shard to make your payment, you'd have to make two different transactions, one on each shard. The client could display this as it would a single transaction, except that you might have something like "65% 3 confirmations, 35% 1 confirmation". In theory there could be cases where the 35% didn't confirm at all. The vendor would need a bit of extra logic to watch for their payment coming through in multiple transactions.

If you needed the whole payment to go through in one transaction - maybe it's a fancy escrow contract or something that needs to be atomic - you'd have to exchange coins on one shard with someone for coins on another shard.

There's no particular reason why people should ascribe different values to different shards - they're all Bitcoins (or whatever alt-coin), they can all be paid to and from the same addresses, and they all look the same in the client. But if you were worried that people might end up valuing different shards differently for some strange reason, you could head the problem off by making a method to move coins across shards. It wouldn't have to be fast or cheap - you'd just need to be able to reassure people that every shard was worth the same, and if anyone got the idea that they were worth different amounts the difference could be arbitraged away.

I don't see how having one currency for 'store of value' and one for 'medium of exchange' is 1) possible to centrally plan, and 2) helps to do anything except dilute mining power and liquidity of both.  Bitcoin is a valuable medium of exchange because it commands significant mining power, it commands significant mining power because the block reward is valuable, and the block reward is valuable because it's a good store of value.  (Yes, 'significant' and 'good' are relative terms.)  And even if the currency functions could be separated this way, how would the 'medium of exchange' currency have avoided the very censorship problem it was designed to solve in the first place for the 'store of value' currency?  It's pretty difficult to move back and forth between the two if either one is censored.

Dubious as it may be FinCEN believes their oversight authority covers the exchanges of virtual currency for other virtual currencies.

yeah, move several BTC (or more!) into LTC/etc altcoin on an exchange.  takes one transaction in the block.  You now can do microtransactions on your altcoin block with no influence on BTC.  when you're done, move LTC/etc altcoin back to BTC on an exchange.  again, one transaction in the block.

And going from virtual currency <-> virtual currency i'm pretty sure was explicitly listed as not subject to regulation.

(which is funny, because it could easily serve as a coin mixer, which i think they tried to call illegal.)

this is precisely why i own some litecoins